Governing R&D Decisions Organisationally
A quantified evaluation process has an organisational precondition: teams holding both authority and resources, functions holding competence but not the casting vote, and senior management gating funding on a shared vision. The prescription dates from 1994, and its reading of prevailing practice belongs to that year.
The sixth dimension is an organisational test, not an analytical one
R9 assesses a decision on six dimensions: frame, alternatives, information, values, logic and commitment. The first five are things an analyst can do at a desk. The sixth is not, and it is set out in full on Decision Quality in R&D.
Commitment asks whether the preceding five were done in a way that generates organisational credibility, acceptance of the results, and commitment to action. That is a question about arrangements — who was in the room, who owns the numbers, who is obliged to act on them — and no amount of analytical rigour answers it.
That coupling is why governance sits on this page and not in a separate discipline. The go/no-go decisions — selection, continuation, modification, termination — are the joint where strategy and execution meet, and they are the decisions the organisational arrangements have to carry.
The organisational prescription, and what each sentence removes
R9's organisational section is three statements long. Read them as a set of transfers rather than as advice, because each one takes something away from somebody.
THE THREE STATEMENTS AND WHAT EACH ONE MOVES
| Statement as given | What it grants | What it removes |
|---|---|---|
| Execute R&D projects through functionally complete teams | A single body that contains every function the project needs — research, development, manufacturing, marketing — rather than a sequence of handovers | The ability of any one function to define the project on its own terms and pass it on |
| Teams are highly empowered by senior management to get the job done | Standing authority to decide within the project rather than case-by-case permission | The escalation habit, and senior management's option to withhold a decision indefinitely |
| Teams must have both authority and resources | Both halves together — authority without resources is a mandate to fail, resources without authority is a budget somebody else spends | The arrangement in which a team is told to deliver and then negotiates for people project by project |
| Functional departments become reservoirs of competence but do not call the shots | A clear and continuing role for the functions: hold, deepen and supply technical capability | Functional veto over project direction, and functional ownership of the schedule |
Paraphrased from the organisational prescription in R9, a 1994 practitioner methods synthesis. The commentary in the third column is drawn out from the statements; the source states the transfers without elaborating them.
Who decides, who analyses, who is accountable
R9 publishes no responsibility matrix. What it does publish is a set of instructions addressed to particular parties, and those instructions can be sorted by who is being told to do something. The table below sorts them, and marks where the source stops.
PARTIES, ASSIGNED WORK AND OPEN QUESTIONS
| Party | What the source assigns it | What the source leaves open |
|---|---|---|
| Senior management | Empowering the teams; requiring an explicit, accepted and shared vision of the product before funding a major new-product development; investigating a portfolio with few high-value projects and many low-value ones | Who inside senior management holds the gate, and what happens when the gate is refused |
| The cross-functional project team | Producing credible assessments of time and cost to completion, probability of success and potential commercial value; explaining how its efforts can be expected to generate value; working the toughest technical hurdles first | Who on the team owns each number, and how disagreement inside the team is resolved |
| Functional departments | Holding and supplying technical competence | How functional technical standards are enforced when the team is the decision-maker |
| R&D management | Monitoring and improving portfolio balance — named as one of its prime responsibilities; asking why any project shows an expected return below one | Whether R&D management can overrule an empowered team, and on what grounds |
| Business management | Reading the probabilistic revenue forecast against business goals and acting when the pipeline will not deliver | The forum in which that reading happens, and its cadence |
Assembled from the decision rules stated in R9. The right-hand column records questions the paper does not answer; it is not a list of the paper's recommendations.
The decision rules the arrangement exists to enforce
Governance arrangements are only worth building if there are rules they are meant to make binding. R9 states its rules in a form that can be tested, which is what makes them governable. The full mechanics of the tools that produce the inputs sit on R&D Project Evaluation Tools.
Rules stated by the source, in conditional form
Why the tools fail without the arrangements
The failure paths are specific, and R9 names most of them itself. None of them is a failure of arithmetic.
Four ways a quantified process dies
Garbage in, garbage out
R9 records that many organisations have had this experience with quantitative R&D evaluation in the past, and positions its toolkit as the fix rather than as self-evidently trustworthy. Estimates produced by whoever is available, unchallenged and unattributed, produce a decision tree that reproduces the opinion it started with.
No credible assessments to decide on
The source states plainly that management rarely has credible assessments on which to base termination decisions, and offers that as part of why uniform cutting persists. The default is not a worse decision method; it is a decision method that requires no evidence.
Analysis nobody owns
The commitment dimension is failed by an evaluation carried out to one side of the people who must act on it. The tools are described as simple to understand but nontrivial to apply, and making them a way of life is said to require sustained effort.
No shared vision at the start
Failure to build a shared product vision from the start is named as one of the most common and most destructive failure modes, causing enormous rework across many industries. It is a framing failure that no downstream valuation detects.
The pattern across all four is the same. The analysis is not wrong; it is unattached. Somebody produced it, nobody is bound by it, and the decision is taken on the grounds it would have been taken on anyway.
Three governance failure modes the portfolio view exposes
Two of these are things organisations do, and one is a thing organisations fail to notice. The portfolio instruments that surface them are described on R&D Portfolio Displays and Strategy Tables.
What the source says organisations do
- Cut every project by the same percentage when budgets tighten. Named as the most common organisational response, and stated to be worse than fully eliminating the weakest projects.
- Present potential revenue as the pipeline forecast — the figure that assumes every project succeeds technically and is commercialised. The source says well-managed R&D organisations do this and lose significant credibility a few years later.
- Defer termination because there is nothing credible to terminate on.
What the source says organisations should do instead
- Eliminate the weakest projects outright unless they can be dramatically improved.
- Report expected revenue — the probability-weighted figure — because potential revenue is described as clearly unrealistic.
- Build the assessments first, so that the continuation and termination conversation has something to sit on.
Reading a 1994 prescription in the present
R9 was published in 1994 in a practitioner journal for R&D and technology managers. Batch rule: where a paper describes tooling, computing, prevailing practice or market conditions, the commentary is dated even where the framework is not. Sorting the two is the reader's job.
WHAT IS DATED AND WHAT IS NOT
| Element | Status | Why |
|---|---|---|
| Functionally complete, empowered teams; functions as reservoirs of competence | Travels | A statement about where authority sits. Independent of technology and of the state of practice in any decade |
| The funding rule, the ranking rule and the gates | Travels | Arithmetic and conditions, both technology-independent |
| The claim that quantitative assessment of cost, time, probability and value was becoming increasingly common for applied R&D | Dated to 1994 | A statement about prevailing practice at the time of writing. It is not evidence about practice now |
| The positioning of decision-quality tools as complementary to, and distinct from, the quality-management toolkit then in vogue | Dated to 1994 | A positioning move against the management fashion of that period |
| The reported benchmarking evidence that leading R&D organisations were using these tools | Externally cited figure, dated 1993 | Quoted by R9 from a study conducted by others, with no percentages given. R9 is not itself evidence for it |
| The anonymised client displays and their values | Illustrative examples | Used to demonstrate the displays. Never benchmarks |
What the source does not settle
One further boundary is explicit. R9 states that none of its discussion is intended to apply to, or constrain, basic knowledge-building research. Governing exploratory science by these rules is applying them outside their stated scope, and the paper says so.
Before you claim a quantified R&D decision process is governed
- Each of the three decision inputs has a named owner who presents it personally
- The funding gate has a named holder, and a refusal at that gate is a real outcome with a known consequence
- The team's authority is written down somewhere other than a slide, and the resources attached to it are committed rather than negotiated per project
- There is a stated route for a decision the team cannot take, and it is used rather than bypassed
- Portfolio balance is reviewed by someone not rewarded on this year's results
- The budget-tightening rule is agreed before the year in which it will be unpopular
- Everyone who must act on the evaluation was involved in producing it
What to carry forward
- Commitment is a dimension of decision quality, not a downstream communication task. An evaluation nobody accepts has failed on the source's own terms.
- The organisational prescription is a transfer of authority: to functionally complete teams holding both authority and resources, away from functions that keep competence but lose the casting vote.
- The rules are governable because they are conditional and testable — fund above the threshold, rank by expected return under constraint, ask why below one, never cut uniformly.
- The failure modes are organisational, not arithmetical: unattributed inputs, no credible assessments to terminate on, analysis nobody owns, and no shared vision at the start.
- Reviewing an R&D function on current results rewards a portfolio loaded with high-probability work and leaves the problem for a successor.
- The source names the end state and not the mechanism. Reporting lines, escalation, cadence and empowerment tests all have to come from somewhere else, and the page you read them on should say so.
Frequently asked questions
Does the source say who should chair the R&D funding decision?
No. It says senior management empowers the teams and should require a shared product vision before funding a major new-product development, but it names no forum, no chair and no cadence. That is one of the clearest gaps in the paper, and it is a gap the reader has to fill from general management practice rather than from this source.
If functional departments do not call the shots, who owns technical standards?
The source does not say. It assigns functions the role of reservoirs of competence and assigns direction to the empowered cross-functional team, but it does not address how a technical standard is upheld against a team decision. Treat this as an unresolved tension rather than a settled arrangement.
Why is cutting every project by the same percentage worse than cutting some entirely?
Because a uniform cut degrades the strong projects to fund the weak ones, and slows everything without removing anything. The source calls the uniform cut the most common organisational response and instructs that the weakest projects be completely eliminated instead, unless they can be dramatically improved.
How can a portfolio look healthy and still be badly governed?
By being loaded with high-probability, lower-value projects. Those deliver on time and flatter current results, while the low-probability work that would sustain the business several years out is missing. The source notes that the manager responsible appears successful for several years and the successor inherits the shortfall.
Is any of this evidence, or is it professional opinion?
It is a consultant's synthesis published in a practitioner journal in 1994, not empirical research. It carries a stated experience base and cites a small number of external studies, but its organisational prescription is a considered position rather than a tested finding, and it should be labelled that way when quoted.
Does it apply to a pure research group?
The source excludes that case explicitly. None of its discussion is intended to apply to, or constrain, basic knowledge-building research; it is written for purposeful work aimed at a commercially attractive objective. Applying the governance rules to exploratory science means applying them outside their stated scope.
References and source attribution
- R9 - improving R&D decisions and execution. Practitioner methods article in a journal for R&D and technology managers, 1994; 8 printed pages; a consultant's synthesis of a decision-analysis toolkit illustrated with anonymised client displays. Not empirical research. Organisational prescription, decision rules and stated limitations drawn on here.
- Externally cited within R9 and not independently verified: a 1993 benchmarking study by a quality-directors network, a decision-quality association and a consulting firm, reported as revealing widespread use of these tools among leading R&D organisations, with no percentages given.
- Eleven copyrighted journal articles on R&D project management, supplied as a reading set for a literature review and profiled for this library. Front matter, abstracts, framework sections, tables and figures were read; article bodies were not reproduced, and all content here is paraphrase. The set was assembled by a student and is not a systematic survey of the field.
- Supplied teaching source for this library (research methods and research process materials). Used here for page conventions and voice only; it does not treat R&D project management.
Suggested questions for Ask KEVOS
- Draft a responsibility split for an R&D funding decision using the roles this source assigns.
- What would I have to change to move my project teams from functional authority to team authority?
- Write the funding gate criteria for a new-product programme, using the rules on this page.
- Test my current R&D portfolio review against the failure modes described here.
- How do I set an agreed budget-tightening rule before the year in which it will be unpopular?
- Which of the gaps on this page would most affect an organisation of my size?